| 1 Dec 2025 |
 raitobezarius | also | 14:45:33 |
| raitobezarius | i wonder | 14:45:34 |
| raitobezarius | if (removeAttrs [ f1_name … fp_name ] A) == (removeAttrs [ f1_name … fp_name ] B) that slow? | 14:45:56 |
| raitobezarius | or some recursiveRemoveAttrs obv | 14:46:05 |
| raitobezarius | in the context of nixpkgs system platform comparison i mean | 14:46:20 |
 piegames | In reply to @raitobezarius:matrix.org
unless we have function semantical equality imho Which, to be clear, we really really don't want, no matter which fancy theoretical models exist that "solve" this problem :) | 15:04:21 |
| raitobezarius | yes yes, we went through this | 15:05:08 |
| raitobezarius | (but u sure u dont want to impl bisimulation?) | 15:05:15 |
| piegames | If the only reason we need function comparison in the first place is "woops we did a Von Neumann mistake and now all our data is mixed with code so we can't compare it anymore", then the solution should be to stop doing that instead | 15:06:00 |
| piegames | In reply to @raitobezarius:matrix.org
(but u sure u dont want to impl bisimulation?) Only on a conference paper | 15:06:19 |
 K900 | Oof | 15:42:00 |
| K900 | Lowdown update in nixpkgs broke the override in lix flkae | 15:42:10 |
|  νεολαμπής [he/him] changed their profile picture. | 16:42:07 |
|  Brisingr changed their display name from Brisingr05 to Brisingr. | 18:40:09 |
|  schromp joined the room. | 20:52:51 |
| 2 Dec 2025 |
 dr_teagle | how would I do "subflakes" mainly for devshells. I have shared dependencies in the top-level flake and then the specific dependencies lower. This is for an obsidian vault | 02:32:59 |
| K900 | Don't | 05:39:21 |
| K900 | Do multiple dev shells in one flake if you must | 05:39:30 |
 just1602 | I've colleague who did that in a couple of repos at work, and it's a mess to work with. You should really listen to k900 advise, IMO | 05:41:14 |
| dr_teagle | understandable | 05:55:16 |
 brendanzab | Hey, just wondering if there is there any consensus yet on what to do r.e. the flake freeze? i.e. https://lix.systems/blog/2025-11-18-lix-2.94-release/#flakes-enters-freeze-period and https://wiki.lix.systems/books/lix-contributors/page/flakes-feature-freeze?
Is it better just to wait and see where things go, trusting that there will be a decent migration period? | 09:44:50 |
| K900 | Don't do anything yet | 09:53:20 |
| K900 | There will be no removal of flakes without some sort of offramp | 09:53:34 |
| brendanzab | Ah, thanks! Was a little unsure from how the announcements were worded. | 09:57:52 |
 helle (just a stray cat girl) | there will also likely be a plug-in or similar situation to allow you to use flakes | 10:27:33 |
 nikstur | Is there a way I can give nix-eval-jobs an --attr flag (just like nix-build) so that it only builds a subset of packages in a file | 10:45:10 |
| nikstur | * Is there a way I can give nix-eval-jobs an --attr flag (just like nix-build) so that it only builds a subset of packages in a file? | 10:45:15 |
|  ghlent joined the room. | 10:56:28 |
| raitobezarius | --expr | 11:27:50 |
| nikstur | Ok so I have to build the expression myself, e.g. --expr '(import ./release.nix).packages' vs --attr packages. | 11:39:46 |
